Add an AutofillProfileComparator class.

components/autofill/core/browser/autofill_profile.cc

 // Copyright 2013 The Chromium Authors. All rights reserved.
 // Use of this source code is governed by a BSD-style license that can be
 // found in the LICENSE file.
 
 #include "components/autofill/core/browser/autofill_profile.h"
 
 #include <algorithm>
 #include <functional>
 #include <map>
 #include <memory>
 #include <ostream>
 #include <set>
 
 #include "base/guid.h"
 #include "base/i18n/case_conversion.h"
 #include "base/i18n/char_iterator.h"
 #include "base/logging.h"
 #include "base/macros.h"
 #include "base/metrics/histogram_macros.h"
 #include "base/sha1.h"
 #include "base/strings/string_util.h"
 #include "base/strings/utf_string_conversion_utils.h"
 #include "base/strings/utf_string_conversions.h"
 #include "components/autofill/core/browser/address.h"
 #include "components/autofill/core/browser/address_i18n.h"
 #include "components/autofill/core/browser/autofill_country.h"
 #include "components/autofill/core/browser/autofill_field.h"
 #include "components/autofill/core/browser/autofill_metrics.h"
+#include "components/autofill/core/browser/autofill_profile_comparator.h"
 #include "components/autofill/core/browser/autofill_type.h"
 #include "components/autofill/core/browser/contact_info.h"
 #include "components/autofill/core/browser/phone_number.h"
 #include "components/autofill/core/browser/phone_number_i18n.h"
 #include "components/autofill/core/browser/state_names.h"
 #include "components/autofill/core/browser/validation.h"
 #include "components/autofill/core/common/autofill_l10n_util.h"
 #include "components/autofill/core/common/form_field_data.h"
 #include "grit/components_strings.h"
 #include "third_party/icu/source/common/unicode/uchar.h"
@@ skipping 172 matching lines @@
         collapsed_set.insert(PHONE_HOME_WHOLE_NUMBER);
         break;
 
       default:
         collapsed_set.insert(it);
     }
   }
   std::swap(*type_set, collapsed_set);
 }
 
-base::string16 NormalizeForComparison(const base::string16& text) {
-  using icu::UnicodeString;
-  using icu::Transliterator;
-
-  // Use ICU transliteration to remove diacritics and fold case.
-  // See http://userguide.icu-project.org/transforms/general
-  UErrorCode status = U_ZERO_ERROR;
-  std::unique_ptr<Transliterator> transliterator(Transliterator::createInstance(
-      "NFD; [:Nonspacing Mark:] Remove; Lower; NFC", UTRANS_FORWARD, status));
-  if (U_FAILURE(status) || transliterator == nullptr) {
-    LOG(ERROR) << "Failed to create ICU Transliterator: "
-               << u_errorName(status);
-    return text;
-  }
-
-  UnicodeString value = UnicodeString(text.data(), text.length());
-  transliterator->transliterate(value);
-
-  return base::string16(value.getBuffer(), value.length());
-}
-
 }  // namespace
 
 AutofillProfile::AutofillProfile(const std::string& guid,
                                  const std::string& origin)
     : AutofillDataModel(guid, origin),
       record_type_(LOCAL_PROFILE),
       phone_number_(this) {
 }
 
 AutofillProfile::AutofillProfile(RecordType type, const std::string& server_id)
@@ skipping 534 matching lines @@
   server_id_ = base::SHA1HashString(contents_utf8);
 }
 
 void AutofillProfile::RecordAndLogUse() {
   UMA_HISTOGRAM_COUNTS_1000("Autofill.DaysSinceLastUse.Profile",
                             (base::Time::Now() - use_date()).InDays());
   RecordUse();
 }
 
 // static
 base::string16 AutofillProfile::CanonicalizeProfileString(

[Mathieu, 2016/06/08 17:56:06]

I think it makes sense to replace uses of CanonicalizeProfileString with

AutofillProfileComparator comparator;

...
(use comparator.NormalizeForComparison(str))

We sometimes use CanonicalizeProfileString on a couple of strings per function
(and in a for loop in one instance). We'll save on the initialization of
AutofillProfileComparator by reusing it!

[Roger McFarlane (Chromium), 2016/06/09 20:19:11]

I'd prefer to do this in a follow-on CL, it touches a lot of code  and prompts a
bunch of other clean-up.

     const base::string16& str) {
-  base::string16 ret;
-  ret.reserve(str.size());
-
-  // This algorithm is not designed to be perfect, we could get arbitrarily
-  // fancy here trying to canonicalize address lines. Instead, this is designed
-  // to handle common cases for all types of data (addresses and names) without
-  // the need of domain-specific logic.
-  //
-  // 1. Convert punctuation to spaces and normalize all whitespace to spaces.
-  //    This will convert "Mid-Island Plz." -> "Mid Island Plz " (the trailing
-  //    space will be trimmed off outside of the end of the loop).
-  //
-  // 2. Collapse consecutive punctuation/whitespace characters to a single
-  //    space. We pretend the string has already started with whitespace in
-  //    order to trim leading spaces.
-  //
-  // 3. Remove diacritics (accents and other non-spacing marks) and perform
-  //    case folding to lower-case.
-
-  bool previous_was_whitespace = true;
-  for (base::i18n::UTF16CharIterator iter(&str); !iter.end(); iter.Advance()) {
-    switch (u_charType(iter.get())) {
-      // Punctuation
-      case U_DASH_PUNCTUATION:
-      case U_START_PUNCTUATION:
-      case U_END_PUNCTUATION:
-      case U_CONNECTOR_PUNCTUATION:
-      case U_OTHER_PUNCTUATION:
-      // Whitespace
-      case U_CONTROL_CHAR:  // To escape the '\n' character.
-      case U_SPACE_SEPARATOR:
-      case U_LINE_SEPARATOR:
-      case U_PARAGRAPH_SEPARATOR:
-        if (!previous_was_whitespace) {
-          ret.push_back(' ');
-          previous_was_whitespace = true;
-        }
-        break;
-
-      default:
-        previous_was_whitespace = false;
-        base::WriteUnicodeCharacter(iter.get(), &ret);
-        break;
-    }
-  }
-
-  // Trim off trailing whitespace if we left one.
-  if (previous_was_whitespace && !ret.empty())
-    ret.resize(ret.size() - 1);
-
-  // Remove diacritics and perform case folding.
-  return NormalizeForComparison(ret);
+  AutofillProfileComparator comparator;
+  return comparator.NormalizeForComparison(str);
 }
 
 // static
 bool AutofillProfile::AreProfileStringsSimilar(const base::string16& a,

[Mathieu, 2016/06/08 17:56:06]

Seems to be unused, remove

[Roger McFarlane (Chromium), 2016/06/09 20:19:11]

Done.

                                                const base::string16& b) {
-  return CanonicalizeProfileString(a) == CanonicalizeProfileString(b);
+  AutofillProfileComparator compare;
+  return compare.NormalizeForComparison(a) == compare.NormalizeForComparison(b);
 }
 
 void AutofillProfile::GetSupportedTypes(
     ServerFieldTypeSet* supported_types) const {
   FormGroupList info = FormGroups();
   for (const auto& it : info) {
     it->GetSupportedTypes(supported_types);
   }
 }
 
@@ skipping 218 matching lines @@
             << " " << UTF16ToUTF8(profile.GetRawInfo(ADDRESS_HOME_CITY)) << " "
             << UTF16ToUTF8(profile.GetRawInfo(ADDRESS_HOME_STATE)) << " "
             << UTF16ToUTF8(profile.GetRawInfo(ADDRESS_HOME_ZIP)) << " "
             << UTF16ToUTF8(profile.GetRawInfo(ADDRESS_HOME_SORTING_CODE)) << " "
             << UTF16ToUTF8(profile.GetRawInfo(ADDRESS_HOME_COUNTRY)) << " "
             << profile.language_code() << " "
             << UTF16ToUTF8(profile.GetRawInfo(PHONE_HOME_WHOLE_NUMBER));
 }
 
 }  // namespace autofill